Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 21924
    Prev
    Next
    INDEX
    Explanations

    phrases indicating the origin or source of content

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "Originally."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     failure
    -0.70
     statistical
    -0.66
     drive
    -0.65
     rout
    -0.63
     fault
    -0.62
     computer
    -0.62
     impl
    -0.61
     pressures
    -0.61
     failures
    -0.61
     repeat
    -0.61
    POSITIVE LOGITS
    Originally
    3.76
     Originally
    3.25
    Initially
    1.49
     Initially
    1.47
    Previously
    1.44
     originally
    1.39
     Previously
    1.35
    Original
    1.22
    orig
    1.17
    Currently
    1.06
    Activations Density 0.016%

    No Known Activations